On Wed, 26 Feb 2025 19:06:08 +0900, Shin'ichiro Kawasaki wrote:
> Jens, please consider to apply this series to v6.15/block.
>
> Currently, null_blk has 'badblocks' parameter to simulate IO failures
> for broken blocks. This helps checking if userland tools can handle IO
> failures. However, this badblocks feature has two differences from the
> IO failures on real storage devices. Firstly, when write operations fail
> for the badblocks, null_blk does not write any data, while real storage
> devices sometimes do partial data write. Secondly, null_blk always make
> write operations fail for the specified badblocks, while real storage
> devices can recover the bad blocks so that next write operations can
> succeed after failure. Hence, real storage devices are required to check
> if userland tools support such partial writes or bad blocks recovery.
>
> [...]
Applied, thanks!
[1/5] null_blk: generate null_blk configfs features string
commit: d8ae0061afb8bbdb0cf6b2cd4b5be5c54e42b228
[2/5] null_blk: introduce badblocks_once parameter
commit: 6b87fa3245a93913efb3d8b858f6750d655d5db9
[3/5] null_blk: replace null_process_cmd() call in null_zone_write()
commit: 7859d042b0954f843d2e97c1324bb04bf28df2f6
[4/5] null_blk: pass transfer size to null_handle_rq()
commit: 6d9725d1000a0bc4e41fbe2db51181e80e4260eb
[5/5] null_blk: do partial IO for bad blocks
commit: 386d7f4be4cdfb0b3a937f26fe674818394f795e
Best regards,
--
Jens Axboe
